Gallium nitride (GaN) LEDs have proven useful for a variety of lighting applications (e.g., full-color displays, traffic lights, etc.), and have potential for even more applications (e.g., backlighting LCD panels, solid state lighting to replace conventional incandescent lamps and fluorescent lights, etc.) if these LEDs can be made more efficient. To realize higher efficiency for GaN LEDs, they need to have enhanced output power, lower turn-on voltage and reduced series resistance. The series resistance in GaN LEDs is closely related to the efficiency of dopant activation, uniformity of current spreading, and ohmic contact formation.
In GaN, a n-type dopant can be readily achieved using Si and with an activation concentration as high as 1×1020 cm−3. The p-type GaN can be obtained by using Mg as the dopant. The efficiency of Mg doping, however, is quite low due to its high thermal activation energy. At room temperature, only a few percent of the incorporated Mg contributes to the free-hole concentration. Mg doping is further complicated during MOCVD growth because of hydrogen passivation during the growth process. Hydrogen passivation requires a thermal annealing step to break the Mg—H bonds and activate the dopant. Typical thermal annealing is performed at about 700° C. in a N2 environment. To date, the practical hole concentration in p-type GaN is still limited to about 5×1017 cm−3. This low activation level leads to poor ohmic contact and a large spreading resistance, which restrict the performance of GaN LEDs.